Discovering territory, object by object
Carpinus is our artificial intelligence technology for automatically segmenting geographic features on high- and very high-resolution orthophotos and satellite images. Thanks to advanced computer vision and deep learning models, Carpinus can automatically and accurately delineate features of interest, transforming pixels into ready-to-use geographic information.
With Carpinus, we can automatically segment buildings, swimming pools, roads, aircraft, ships, and other features of territorial interest, which reduces manual photointerpretation tasks, speeds up map updating processes, and cuts down on field and office inspection time, thereby lowering costs.
Carpinus‘ main strength lies in its ability to compare observed reality with information recorded in official databases, such as cadastral records. It automatically detects discrepancies, changes, and new incidents on the ground. This capability is particularly valuable for agencies responsible for cadastral management, urban planning, land control, and updating geographic inventories.
Carpinus isn’t just a segmentation technology, however. It’s a tool designed to turn images into decisions, from spotting cadastral omissions to locating illegal construction.
Carpinus use cases
The segmentation of building in urban areas
This PNOA image of San Vicente de la Barquera (Cantabria), with a resolution of 15 cm, shows how Carpinus accurately segments all buildings located in urban areas.
PNOA
PNOA con Carpinus
The segmentation of buildings in rural areas
This PNOA image of O Reguerio (A Coruña), with a resolution of 15 cm, shows how Carpinus accurately segments all buildings in a rural area, including a granary (marked in red).
PNOA
PNOA con Carpinus
The segmentation of roads
In this Sentinel-2 image covering the Leganés, La Fortuna and Villaverde area (Madrid), Carpinus segments roads, streets, interchanges and roundabouts (in red), thereby generatin up-to-date automated mapping.
Sentinel-2
Sentinel-2 con Carpinus
The segmentación of swimming pools
In this 50 cm resolution Pléiades image of a location in the province of Pontevedra, can see how Carpinus precisely segments the swimming pools found in the built-up area. Pools that are full and clean are shown in blue, those that are full and dirty in green, empty pools in yellow and covered pools in orange.
Pléiades
Pléiades con Carpinus
Detection of cadastral omissions
Using Carpinus technology, we can also detect changes in segmented buildings by comparing this information with official cadastal data. In this example, obtained from the province of Cádiz, the yellow highlights indicate newly segmented structures that do not appear in the official cadastral mapping.
PNOA con Catastro
PNOA con Carpinus y comparativa con Catastro
